Google AI can now pull automatically created assets (ACA) based on the context of a search query itself.
Google announced this new ACA program feature at Google Marketing Live today.
Until now, Google Ads could skim landing pages using AI and machine learning to automatically create assets from various eligible sources, based on predictive performance and relevance. But this addition of generative AI offers a new twist for advertisers.
Why we care. The addition of generative AI addition to ACAs should be a boon to relevancy for advertisers. Leveraging other ads and landing pages for better query matching should make ads more dynamic and effective if you are ok with giving up control to Google AI.
But be careful. While this technology can greatly improve your ad’s relevancy, it might not be the best bet for highly regulated industries or brands with stringent compliance standards. By giving up complete control to AI, Google will be the final say on what is displayed on your ads. Advertisements are regulated by the FTC and you won’t be able to guarantee the message that will be displayed by the ACAs.
What’s different? In the past, Google would use “eligible sources” to power the ACAs before heading into the auction. Now, Google AI will process the context of a query to adapt search ads toward user intent.
For example, Google showed the search query [skin care for dry sensitive skin]. Google’s AI could use content from your landing page and existing ads to generate a targeted headline titled “Soothe Your Dry, Sensitive Skin.”
- Another way to think about this: the AI-powered ACAs are a more robust and modern version of Dynamic Keyword Insertion for advertisers.
Campaign level. Generative AI for ACAs will be implemented at the campaign level, not the account level, which will lend itself nicely to testing as you can control what campaigns are powered by the AI.
Availability. This is in a global open beta for the English language, with plans to expand to additional languages this year.
Opinions expressed in this article are those of the guest author and not necessarily Search Engine Land. Staff authors are listed here.
